© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 1536
    9. 8-RES_FS1536-JB
    11. 49
    Prev
    Next
    INDEX
    Explanations

    information related to news articles and legal issues

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs1536-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_1536
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    1,536
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    373
    +0.13
    0.6%
    640
    +0.12
    0.6%
    372
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    640
    +0.13
    0.24
    6
    +0.12
    0.17
    94
    +0.11
    0.16
    Negative Logits
    lihood
    -0.85
    ignty
    -0.83
    CONCLUS
    -0.71
    yss
    -0.68
     someday
    -0.65
    gren
    -0.65
    istical
    -0.64
     regulates
    -0.62
    aterasu
    -0.62
    uph
    -0.62
    POSITIVE LOGITS
    Interested
    1.20
    WASHINGTON
    0.99
    Specifically
    0.99
    Speaking
    0.93
    According
    0.91
    SHARE
    0.87
    Sources
    0.86
    SEE
    0.85
    article
    0.85
    Writing
    0.84
    Activations Density 1.963%

    No Known Activations